HackerTrans
TopNewTrendsCommentsPastAskShowJobs

tevlon

320 karmajoined vor 11 Jahren

Submissions

We Must Act Now – A Statement on AI's Transformation of the Economy

wemustactnow.ai
9 points·by tevlon·vor 16 Stunden·2 comments

[untitled]

1 points·by tevlon·vor 2 Monaten·0 comments

Show HN: Xona.ai 2.0 – create beautiful interiors in seconds

xona.ai
7 points·by tevlon·vor 9 Monaten·2 comments

I solved almost all of free problems on LeetCode using AI

old.reddit.com
3 points·by tevlon·letztes Jahr·5 comments

Show HN: Anschreibenai.com – an AI coverletter generator. it's free. no account

anschreibenai.com
3 points·by tevlon·letztes Jahr·0 comments

comments

tevlon
·letzten Monat·discuss
Couple days ago, i used Claude to implement an improved version of gpt-1. I am no ML Engineer by no means. I am just a normal backend engineer. I ended up creating a hybrid between gpt-1 and modded-nanogpt (from KellerJordan).

I was able to reproduce the results of the original gpt-1 paper with my gaming PC. I don't even have alot of VRAM. My NVIDIA GeForce RTX 2060 SUPER was able to reproduce most of the results with just 1 hour of training. I would totally recommend to do the same, if you are interested in pre-training LLMs.

The code is here: https://github.com/epoyraz/modded-gpt-1 But, you can also just ask Claude 4.8 or Codex 5.5
tevlon
·vor 9 Monaten·discuss
Thank you for your kind words. Yes, i hate subscriptions. But, also. Renovating is a one-off thing and i can't justify charging a monthly fee for a service that is only needed once.

This is one of the reasons people love using xona: no subscriptions :)
tevlon
·letztes Jahr·discuss
That's why i use Docker MCP Catalog. One MCP Server to rule them all. more info: https://www.youtube.com/watch?v=6I2L4U7Xq6g
tevlon
·letztes Jahr·discuss
makes sense
tevlon
·letztes Jahr·discuss
i was thinking the same. And i am sure it is in the training set, but i couldn't solve 10 hard problems. Also, most of the hard problems, the LLMs weren't able to solve in one shot. How do you explain this?